Dual Quaternion based Compliant Movement Primitives for Deformable Object Manipulation
Amir SAMAI, John Thomas, Mohammad Alkhatib, Erol Ozgur, Youcef Mezouar
AI summary
Problem
Existing learning-from-demonstration methods for deformable objects either ignore critical interaction forces or rely on robot-specific joint-space formulations, limiting safety, compliance, and cross-platform transferability.
Approach
The method encodes 6-DoF end-effector motion and interaction forces in operational space using dual-quaternion Dynamic Movement Primitives and learnable wrench primitives, generalized to new starting poses via manifold-aware Gaussian process regression.
Key results
- Compliant execution of contact-rich tasks like shoe-sole detachment and foam bending
- Accurate generalization of learned wrenches to unseen initial poses via SE(3)-aware Gaussian process regression
- Singularity-free, robot-agnostic 6-DoF pose and force encoding in operational space
- Synchronized reproduction of motion and force profiles directly at the end-effector
Why it matters
Provides a model-free, transferable framework for safe and compliant deformable-object manipulation across different robots and task variations.
Abstract
Learning from demonstration effectively transfers human manipulation skills to robots. It can be especially useful for imitating industrial manipulation tasks which are performed by humans and are difficult to model such as deformable object manipulation. Manipulation of deformable objects often requires not only accurate tracking of the demonstration tra- jectory using a robot end-effector, but also the accommodation of interaction forces. Precise tracking of such trajectories while ignoring these interaction forces leads to overly stiff, unsafe, or unsuccessful executions. We address this problem by proposing Dual Quaternion based Compliant Movement Primitives (DQ- CMP). DQ-CMP couples a dual-quaternion based Dynamic Movement Primitive for compact 6-DoF pose encoding with learnable wrench primitives. This combination reproduces synchronized motion and force behaviors directly at the end effector. The method is robot-agnostic and singularity-free at the representation-level, as it operates in operational space using dual quaternions. From a few demonstrations, required wrenches for unseen initial configurations are predicted using Gaussian process regression defined on the pose manifold. This enables generalization of the learned wrenches across different starting poses. We validate the method on real-robot experiments including a shoe-sole detachment for recycling and bending of stiff foam inside a box. Results show compliant, safe task execution and successful generalization to new initial poses.